This page gives an overview of the Tools API which provides turnkey text and sentiment analysis functionality for understanding opinionated content such as user reviews, customer feedback, etc.

🚧

API Key

To use the Tools API, you must have an API key. Contact [email protected] to obtain a key.

The Tools API provides the following functionality:

  • Sentiment analysis - extracts opinions from text, optionally identifying key topics for each opinion
  • Topic extraction - identifies important entities (brands, names, etc.), features, and descriptors in a piece of text
  • Summarization - distills content into overall sentiment, representative quotes, and best and worst features
  • Feature sets - allows storage of taxonomies to be reused in sentiment analysis

Analysis endpoints handle POST requests with parameters supplied via JSON objects in the message body.
$obj = array(
	"text" => "I love natural language processing!  NLP is lots of fun.",
	"entities" => array(array("nlp", "natural language processing"))
);

$options = array(
  'http' => array(
    'method'  => 'POST',
    'content' => json_encode($obj),
    'header'=>  "Content-Type: application/json\r\n" .
                "Accept: application/json\r\n"
    )
);

$context  = stream_context_create($options);
$result = file_get_contents("http://api.addstructure.com/v1/analysis/sentiment?key=APIKEY", false, $context);
echo($result);

The output is also formatted as JSON.

{
  "success":true,
  "message":"",
  "opinions":[
    {"sentenceID":"s1", "entity":"nlp", "feature":"general", "sentiment":1, "emotion":"joy", "sentence":"I love natural language processing!", "segment":"I love natural language processing!", "showsIntent":false},
    {"sentenceID":"s2", "entity":"nlp", "feature":"general", "sentiment":1, "emotion":"none", "sentence":"NLP is lots of fun."," segment":"NLP is lots of fun.","showsIntent":false}
  ]
}